COPYRIGHT NOTICE Zhiru/The Making of a Savior Bodhisattva is published by University of Hawaii Press and copyrighted, 2007, by the Kuroda

a Institute. All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher, except for reading and browsing via the World Wide Web. Users are not permitted to mount this le on any network servers.

INTRODUCTION

Problems and Perspectives

My rst encounter with the Bodhisattva Dizang (Kitigarbha) took place more than two decades ago at the Chinese temple of a lay Buddhist society in Singapore called the Buddhist Lodge (Jushi lin ). It was the last night of the seventh lunar month of the Chinese calendar, a month traditionally consecrated to the welfare of deceased relatives, especially those reborn in the unfortunate realms of hungry ghosts and hell beings.1 An elaborate festival brought to a close the month-long communal recitation of the Dizang pusa benyuan jing (Scripture on the Past 2 Vows of Dizang Bodhisattva). At the heart of this festival was a ritual dramatization of the feeding the hungry ghosts ceremony (shishi ).3 The monk who headed the performance wore ne, colorful robes and a golden ve-buddha crown to signify his sovereignty over the infernal realm. A small statue of Dizang Bodhisattva sat on the ritual table.4 Uttering a sequence of dhras and liturgical prayers, the monk sprinkled water on the
1. The thirtieth day of the seventh lunar month is celebrated as the day when Dizang Bodhisattva achieved awakening during his incarnation as the Silla monk the Golden Dizang (Jin Dizang ). Jin is said to have resided at Mount Jiuhua (Jiuhua shan ), now famed as the pilgrimage site for the Bodhisattva Dizang. Throughout the seventh lunar month rituals are held at Mount Jiuhua, culminating in mass ceremonies on the last day of the month. Devotees and pilgrims ock to the mountain to participate in the nal rites. Local temples and monasteries around China, as well as immigrant societies outside of the country, also observe the rituals of the seventh lunar month honoring Dizang Bodhisattva. 2. Dizang pusa benyuan jing, T412:13.777c790a, attributed to iknanda (Shichanantuo , 652710). Today Chinese Buddhists frequently employ this scripture at Dizang ritual assemblies. However, in premodern times, a small body of Dizang liturgical works was used in conjunction with confession rites addressed to Dizang Bodhisattva. For a discussion of these premodern repentance rituals, see Wang 1999. 3. This is basically the ritual performance of an esoteric ceremony introduced in the Tang period (618907). For studies on shishi, see Orzech 1994; Lye 2003. The earliest versions in the corpus of ritual texts on feeding the hungry ghosts as it exists today date to the Yuan period (12801368). However, the ritual can be traced to esoteric Buddhist practices of the eighth century. For a translation of the portion of the ritual that can be traced to the Tang period, see Orzech 1996; for a translation of a version dating to the Yuan, see Stevenson 2004. 4. The ritual specialist wears the red and yellow ritual robes of an abbot. On the symbolism of monastic robes, see Kieschnick 1999. At the Buddhist Lodge, a statue of Dizang is always placed on the ritual table. (However, a statue of Guanyin may also be used.) One ritualist I interviewed informed me that the statue represents the principal deity (benzun ) for the ritual, which is determined by the choice of deity for the visualization. However, as I will clarify later, in the ritual text Guanyin is the principal deity.

Introduction

ground to sanctify the ritual space then symbolically enacted the closing of the gates of hell. Engulfed in incense smoke, candlelight, and mesmerizing incantations, the room was dramatically transformed into the desolate underworld of the damned. Excitement ran through the crowd of lay participants and observers as they stood on their toes and craned their necks to catch sight of the ritualists every hand gesture (Ch. shouyin ; Skt. mudr). I heard it murmured that he was impersonating Dizang returning the spirits of the deceased to their subterranean connes after a month in the world of the living. This then was my introduction to Dizang Bodhisattvaan impressive, shaman-like gure who oversees the affairs of the dead, an afterlife deity inextricably bound to the terrifying ethos of the underworld, a symbol of light and salvation for the lost and condemned souls of the deceased. Years later I realized that the ritual I had witnessed was a modied enactment of a long ritual text, the Yuqie jiyao jiu Anan tuoluoni yankou guiyi jing (The Scripture from the Essentials of Yoga Teachings on the Dhra Flaming-Mouth Rites for Saving nanda), ascribed to the esoteric Buddhist teacher Amoghavajra (Bukong jingang , 705774).5 This eighth-century text represents an esoteric articulation of a Buddhist rite conjured up to accommodate the indigenous Chinese practice of presenting offerings to honor ancestors. It is commonly known as the liberating the aming-mouths (fang yankou ), or feeding the hungry ghosts, ceremony. According to Amoghavajras text, the purgatories are crushed, the hungry ghosts summoned, and their transgressions evaluated. Following their confession and redemption, the ghosts receive sweet nectar, and the scorching ames in their throats are quenched. In the ritual text the main deity is Guanyin (Avalokitevara), not Dizang. It contains only a short segment where the ritualist, having rst visualized himself as Guanyin, briey transforms into Dizang.6 The theme of feeding hungry ghosts (frequently understood as ancestral ghosts) is present in Tang (618907) and Song (9601279) afterlife rituals like the Avalambana, or Ullambana (Yulanpen ), a ceremonial offering to the monastic establishment on the behalf of ones deceased ancestors; the Daoist version of the Avalambana, the Middle Primordial (zhongyuan ) festival; and the Buddho-Daoist purgatorial rites known as Universal Salvation (pudu ).7 In a popular temple setting like the Buddhist Lodge,
5. Yuqie jiyao jiu Anan tuoluoni yankou guiyi jing, T1318:21.468c472b; cf. Yuqie jiyao yankou shishi yi , T1320:21.473c484a. Also see n. 3. 6. Hun Yun Lye rst brought to my attention the lack of textual support for the relationship between the ritual and Dizang Bodhisattva. The visualization procedure is more clearly delineated in the Yuqie jiyao yankou shishi yi, where the ritualist rst envisions Guanyin (T1320:21.476bc), then, having assumed the form of Guanyin, visualizes Dizang, who has aspired to release all beings from hell (T1320:21.476c). 7. The Sanskrit word avalambana (ullambana in Pli) literally means hanging upside down and is associated with a physically afictive condition in hell. The Chinese term yulan-

Problems and Perspectives

these various rites of feeding and liberating ancestral ghosts are drawn into the nal rites of fang yankou, over which is transposed the imagery of Dizang Bodhisattva, who is invoked especially in scriptural and liturgical recitations throughout the seventh lunar month. In temple observances of holy days, the thirtieth day of the seventh lunar month is today celebrated as Dizang Bodhisattvas birthday.8 Lay Buddhists make monetary donations to have the names of their departed loved ones written on pieces of paper, which are then pasted on makeshift memorial tablets housed in tents erected on the temple grounds during the seventh month.9 Upon completion of the rites on the last day of the month, the paper tablets are burned in a bonre, and the lay sponsors return to their homes satised with the knowledge that they have taken care of their deceased family members. Dizang is best known in Chinese religion today as the savior par excellence of the dead, especially of those undergoing torments in hell because of their wicked ways. In Chinese monasteries and temples, Dizang Bodhisattva is frequently enshrined separatelyaway from the main shrine hallin an adjunct Merit Cloister (Gongde yuan ), which houses the memorial tablets of the dead ancestors and relatives of lay patrons. Dizang is often portrayed as a crowned sovereign, wearing monastic robes and a ve-petaled golden crown (see Figure 1). He is also represented as a golden-bodied monk with a shaven head dressed in somber robes, his face serene and eyes downcast in silent contemplation (see Figures 2 and 3).10 In both cases Dizang is usually shown holding a staff in one hand and a jewel in the other, a conguration of attributes connected to his role as Lord
pen is a euphonic transliteration of the Sanskirt words and does not quite communicate the semantics of the Indian phrase. For a study in English of the Ghost Festival in medieval China, see Teiser 1988b; also see Teiser 1986. The Avalambana focuses on ceremonial offerings to the sagha after the rain retreat. Especially in China, the ceremony focuses on transferring merit to dead ancestral ghosts who can partake of food offerings as a result of merit from making offerings to the monastic community. The mythological foundation for the Ghost Festival is based on the legend of Maudgalyyana (Ch. Mulian ; J. Mokuren), who through the Buddhas aid and by virtue of offerings made to the monastic community, saved his deceased mother from the sufferings of hell. For a translation of one Japanese version of the legend and a discussion of the Ghost Festival in Japan known as urabon, see Glassman 1999. On the fteenth day of the seventh lunar month, Daoists hold a festival similar to the Ullambana known as zhongyuan, designating the day on which the Middle Primordial (that is, zhongyuan) descends to earth to judge the deeds of the people. Offerings are made to gods and ancestors by way of the Daoist temple and its priests, and consequently hungry ghosts are able to eat their ll on this day. For a discussion of the medieval sources on the zhongyuan festival, see Teiser 1988b: 3540. On Daoist purgatorial rites in connection with a twelfth-century Daoist visualization text that discusses the interrelation between Buddhist and Daoist expressions of the rites of feeding the hungry ghosts, see Boltz 1983 (esp. 508509). For a comparative study of fang yankou and pudu, see Orzech 2002. 8. For a listing of buddha and bodhisattva birthdays, see Zhong 2000: 186. 9. This practice is commonly known as chaodu , or to bring [the deceased] over [to better states of rebirth]. 10. Dizang is rst depicted as a monk in the Shilun jing , a text discussed in Chapter 1.

Figure 1. Crowned Dizang. Tiantai shan Fangguang si, Zhejiang.

Figure 2. Dizang, modern statue.

Problems and Perspectives

Figure 3. Dizang, modern statue. Buddhist Lodge Memorial Hall, Singapore.

Teacher of the Desolate Darkness (youming jiaozhu Hymn to Dizang (Dizang zan ) states:

). Hence, the

His resplendent jewel illumines completely the road to heaven; His golden staff quakes open the gates of hell. He welcomes and leads inexhaustible generations of families and relatives; And on the bank of [the lake of] the throne of the nine [grades of] lotus,11 they prostrate themselves to the compassionate lord.

12

Given Dizangs dominance in the cult of the dead in modern Chinese religion, it is only natural that participants and observers alike equate Dizang with the ritualist in his monastic robes and crown enacting the scene of closing the gates of hell. He is, in short, the sovereign of the underworld.13
11. The nine grades of lotus occur in the Foshuo guan wuliangshou fo jing (also known as the Guan jing), T365:12.344c346a. The Guan jing enumerates sixteen visualizations that a practitioner should observe to attain rebirth in Amitbhas Pure Land in the west. In visualizations 1316, the practitioner envisions himself being reborn on a different grade of lotus (superior, middling, or inferior of the superior grade; superior, middling, or inferior of the middling grade; superior, middling, or inferior of the inferior grade). 12. The Hymn to Dizang is recited during liturgies addressed to Dizang Bodhisattva. 13. In oral interviews, one ritualist at a rural temple in Xinzhu (Taiwan), an expert in performing the fang yankou, informed me that it is possible to visualize Dizang or Guan-

Introduction

In present-day Chinese Buddhism, Dizang is moreover known as one of the four great bodhisattvas (si da pusa ), together with Guanyin, Puxian (Samantabhadra), and Wenshu (Majur). Each bodhisattva presides over one of four mountain cults, which constitute the tradition of Buddhist pilgrimages hailed as the four famous great mountains (si da mingshan ).14 Dizang Bodhisattva reigns over the Mountain of Nine Flowers (Jiuhua shan ), where he supposedly resided in the late eighth century in his earthly incarnation as a Korean prince-turned-monk known as the Golden Dizang (Ch. Jin Dizang ; Kor. Kim Chijang).15 The pilgrimage cult of Mount Jiuhua and its mythology are today considered central to the Dizang cult. The iconography, myths, sites, rituals, and roles associated with Dizang as they are known today developed over several centuries; they do not always render the complex and shifting vicissitudes this deity has undergone, especially during the medieval period. For instance, Mount Jiuhua emerged visibly as the cultus locus of Dizang devotion only in the late imperial period, long after Dizangs introduction into Chinese society no later than the sixth century.16 It is thus important to study the emergence of the Dizang cult in its historical and social reality to recomplexify Dizang and so help to restore to life (as much as possible from the surviving fragmented historical sources) what this gure meant to Chinese Buddhists from the sixth to tenth centuries.

Unveiling the Presuppositions

Modern scholarly perceptions of Dizang Bodhisattva are shaped largely by two sets of assumptions, each linked to a methodological presupposition in the study of Chinese Buddhism. On the one hand, scholars tend to contrast the early obscurity of this bodhisattva in India, where he was known as Kitigarbha, with his subsequent ascendancy in China as a major Buddhist deity commanding widespread veneration. This contrast of roles in the two geographical regions leads to the assumption that a radical transformation must have taken place during the introduction of Kitigarbha as Dizang to China. For scholars, the radical changes that this bodhisattva seems to have undergone in China make him a natural candidate for studying patterns of sinicization in Chinese Buddhism. The history of Dizang is then a tale of how and why he red the Chinese imagination, and the reconstruction of
yin as the principal deity. The choice of deity is decided by the daochang , or ritual sanctuary, that is, the monastery or temple hosting the ritual. 14. The four pilgrimage centers are: for Guanyin, Putuo shan (Zhejiang); for Dizang, Jiuhua shan (Anhui); for Puxian, Emei shan (Sichuan); for Wenshu, Wutai shan (Shanxi). For a study of the four mountain sites in Chinese Buddhism, see Kamata 1987. The phrase four famous great mountains evidently was used only during the reign of the Qing emperor Kangxi (16621722); see Pan 2000: 820. For further discussion on the emergence of the tradition of the four famous great mountains, see the Conclusion. 15. For studies on Mount Jiuhua, see Powell 1987, 1993; Wang-Toutain 2001. 16. See the Conclusion for further discussion on Mount Jiuhua and Dizang worship.

Problems and Perspectives

his cult becomes an opportunity to demonstrate the Buddhist assimilation of indigenous Chinese elementsjust as scholars have shown Guanyin to be a Chinese transformation of the Indian Avalokitevara.17 In other words, Dizang is regarded as the Chinese bodhisattva par excellence, and the study of Dizang is inevitably linked to the model of Buddhist sinicization that has characterized much of modern scholarship on Chinese Buddhist history. Studies of Dizang also tend to emphasize his role as Supreme Lord of the Underworld, the function for which he is most widely known today. Scholars have for the most part accepted this unidimensional portrayal of Dizang as true not only in the modern period, but also earlier. Consequently, the study of Dizangs history is presented largely as tracing a trajectory culminating in his consecration as the Buddhist sovereign of the underworlda process that necessarily casts Dizang a product of folklorization or vulgarization (minjian hua ) through which Buddhism inltrated into the heart of everyday religion in medieval Chinese society. Viewed through this lens, Dizang has all but lost the character of a bodhisattva. Instead he has joined the complex bureaucracy of Chinese afterlife deities who maintain the records of life and death, judge the actions of the deceased, and pass sentence on their next rebirth. Furthermore, scholars tend to synthesize the two sets of assumptions and so regard the culmination and success of the sinicization process to be the transformation of the Indian Kitigarbha Bodhisattva into the Chinese Dizang, Lord of the Underworld, and his penetration into the everyday religion of medieval Chinese society, which emphasizes the family structure in religion, particularly the obligations of descendants to procure salvation for deceased ancestors and relatives and to pacify those restless spirits threatening domestic bliss. In other words, Dizang, as Lord of the Underworld, is considered a paradigmatic sinicized bodhisattva. Certainly, broaching the study of Dizang through questions of sinicization and folklorization of Buddhism in China does offer salient insights into the rise of Dizang worship. But because these approaches exclude other roles Dizang may have played in medieval Chinese religion, they in fact restrict our understanding of the Dizang cult in China.

A Sinicized Bodhisattva?
First, let us examine the assumption that this bodhisattva underwent a radical transformation after his introduction to China and that he should be
17. The sinicization of the bodhisattva ideal has been discussed in past scholarship. For example, Robert Gimello (1978) examined the early medieval Chinese amalgamation of the Confucian sage and the bodhisattva ideal through gures like Vimalakrti. Jan Yn-hua (1981) studied the sinicization of the bodhisattva in Chinese Buddhist literature through its typology and signicance. Lewis Lancaster (1981) analyzed the bodhisattva doctrine as presented in the Chinese Buddhist canon. Studies have also focused on specic bodhisattvas, especially the savior bodhisattvas. A recent example (2001) is Y Chn-fangs Kuan-yin: The Chinese Transformation of Avalokitevara. Also see Y 1992, 1994.

Introduction

regarded as a sinicized bodhisattva. When tracing his origin to pre-Chinese sources, one immediately encounters an anomaly: The earliest evidence for Kitigarbha in India dates no earlier than the eighth century, by which time, as documented in textual and visual materials, Dizang worship was already present in China, replete with iconography, mythology, texts, and practices. In fact, the earliest scripture on this bodhisattva is a Chinese text, the Shilun jing , or the Daacakra-stra (Scripture on the Ten Wheels), purported to be an anonymous translation from the Northern Liang (397439) but only dated rmly to the sixth century. (See Appendix 1 for arguments on the texts dating.) In addition to the paucity of Indian sources, scholars have also called attention to the fact that, in travelogues of the sixth and seventh centuries, Chinese pilgrims make no mention of the worship of Kitigarbha at the Indian sites they visited.18 In the pantheon of Indian Mahyna, moreover, little is known about Kitigarbha beyond his appearance as one of the eight bodhisattvas. Hence, scholars have concluded that Kitigarbha was either a minor gure or had only a nominal existence in Indian Buddhism. From this perspective, the Scripture on the Ten Wheels is accepted as a plausible pre-Chinese source for Kitigarbha that functioned as the substratum for the sinicization processes that produced the Bodhisattva Dizang. Another hypothesis argues that Dizang originally existed as an audience bodhisattvathat is, a bystander present at a buddhas sermonin the long directories of bodhisattva names preserved in Mahyna scriptures.19 Shortly after his introduction in China in the fourth century, Chinese Buddhists appropriated the name and eshed out Dizangs character so that he became a major object of Buddhist worship, second in popularity only to Guanyin. The origin of the Scripture on the Ten Wheels remains uncertain. The Tibetan version was translated from a Chinese text, so one cannot completely rule out the possibility that this scripture was composed in China.20 From such a perspective, Dizang is better understood as a Chinese bodhisattva who, except perhaps for his name, was engendered almost entirely in Chinese culture and society. Thus it is assumed that the character of this bodhisattva, as well as the iconography, mythology, rituals, and texts of his cult, was largely elaborated in China. Both of these hypotheses accentuate sinicization as the process that created the Bodhisattva Dizang we know today. Yet examining Dizang through the lens of Buddhist sinicization locks us into a stalemate because his rela18. Both the Datang xiyu ji , by the famous Chinese pilgrim-monk Xuanzang (602664), and the Nanhai jigui neifa zhuan , by Yijing (634713), contain no mention of Kitigarbha worship. 19. Lancaster 1981: 155156. For a discussion of examples, see Appendix 2. 20. In Sanskrit, the title is reconstructed as *Daacakrakitigarbha-stra. A colophon appended to the extant Tibetan translation announces it to be based on a Chinese original. The Tibetan title is given as Dus-pa chen-po las Sai snyning-poi khor-lo bcu-pa zhes-bya-ba then-pa chen-poi mdo. The colophon mentions that the text was translated from the Chinese by Hoshang zab-mo, Rnam-par mi-tog, and others.

Problems and Perspectives

tionship to his Indian counterpart, Kitigarbha, is fraught with ambiguity (see Appendix 2). The iksammucaya, attributed to the eighth-century Nalanda monk ntideva, cites passages from one Aryakitigarbha-stra, which was evidently some version of the Scripture on the Ten Wheels.21 ntideva knew of Kitigarbha Bodhisattva and he explicitly invokes Kitigarbha as a member in the retinue of great bodhisattvas.22 By the eighth century in India, Kitigarbha was incorporated into the cult of the eight great bodhisattvas (Ch. ba da pusa ; Skt. Aamahbodhisattva), which was in turn integrated into esoteric Buddhism. However, the paucity of dated evidence means that the process by which Kitigarbha became one of these eight bodhisattvas is lost to us. Some scholars suggest that this bodhisattva may have originated in Central Asia, but again the evidence is hardly conclusive. There is a dearth of datable evidence of Kitigarbha in Central Asia, and the earliest materials date no earlier than the eighth century.23 The early history of Kitigarbha in India and Central Asia remains largely unknown to us, so the task of ascertaining what elements of the Dizang cult were originally derived from India and Central Asia is necessarily doomed from the beginning. Thus it is impossible to determine conclusively whether Dizang is a sinicized bodhisattva.

Sinicization Problematized
In addition to the problems of the evidence, there are also the methodological problems associated with studying Chinese Buddhist history through the lens of sinicization. The traditional approach to sinicization can be polarized into the transformation and conquest models, represented respectively by Erik Zrchers The Buddhist Conquest of China and Kenneth Chens The Chinese Transformation of Buddhism.24 The conquest model argues that from the beginning of the common era, Buddhism began to inltrate China
21. The Aryakitigarbha-stra is no longer extant. The Chinese text obviously predates the iksammucaya. 22. ntideva invokes kagarbha and Kitigarbha, together with Samantabhadra, Majughoa, and Avalokita, as the great bodhisattvas; see Bodhicaryvatra 3.52, translated in Crosby and Skilton 1996: 18. 23. One argument for the Central Asian origin is the attribution of the Benyuan jing to the Khotanese monk iknanda. However, because the textual history of the Benyuan jing is itself problematic, it offers little conrmation of the Central Asian origin for Kitigarbha. See the discussion on the Central Asian sources in Appendix 1. 24. Zrcher 1972; Chen 1973. Several studies have examined various aspects of Chinese Buddhist history through the perspective of Buddhist sinicization in China. An early example is by Tso Sze-bong (1982), who traced the Chinese transformation of the vinaya. Peter Gregory (1991) investigated the sinicization of Buddhist doctrine in the major Huayan thinker Zongmi (780841). In a dissertation study, Kyoko Tokuno (1994) explored the patterns of sinicization that produced the indigenous Chinese scripture Tiwei boli jing . Kuo Li-ying (1994a) discussed patterns of sinicization in Buddhist confession and contrition rites from the fth to the tenth centuries. As mentioned earlier, Y Chn-fangs recent study (2001) examined the transformation of Guanyin in China.

10

Introduction

so pervasively that what was originally a foreign religion became part and parcel of Chinese culture and society. The transformation model asserts that indigenous culture and institutions so radically changed Buddhist teachings and practices in China that Chinese Buddhism must be recognized as a product of sinitic civilization. Robert Sharf has challenged both of these approaches, which, as he accurately points out, have cast the study of Chinese Buddhism permanently in the shadow of Indian Buddhism so that Chinese Buddhist expressions are always evaluated against some pristine Indian original. Sharf argues that the Buddhism medieval Chinese encountered in the translated texts was already sinied if only by virtue of being rendered . . . into the native tongue.25 The Chinese, moreover, approached translations of Buddhist texts not as glosses on the Indic originals, but as valuable resources that addressed their own immediate conceptual, social, and existential concerns. 26 When Chinese Buddhism is understood in these terms, knowing the Indian antecedents proves far less urgent than knowing the questions Chinese readers of Buddhist texts were askingquestions whose historical, linguistic, and conceptual genealogy was largely Chinese. 27 At rst glance, my study seems to adopt Sharfs perspective as its framework because it opens directly with Dizangs manifestations in Chinese contexts and explores this bodhisattva cult from the perspective of medieval Chinese culture, religion, and society. I want to clarify at the outset that this study does not share Sharfs presumptions or conclusions about the nature of Chinese Buddhism. Sharfs perspective has salient meritsnot the least of which is questioning the scholarly use of Indian Buddhism as the authoritative standard for evaluating Buddhist developments in China as well as reiterating the urgency to study and contextualize Chinese Buddhism within the local social and institutional structures and what he calls the local episteme. 28 Nonetheless, brought to its logical implication, Sharfs polemical arguments run the risk of implying a Chinese Buddhism that is culturally self-contained, impermeable, and relatively isolated from other Buddhist geographical regions. This is surely a problematic reductionist rendition because it belies the regular contacts and mutual inuences across the land and sea routes that continued throughout the history of Chinese Buddhism, long after the initial transmission.29 Medieval Buddhists in China for the
25. Sharf 2002a: 19. 26. Ibid.: 12. 27. Ibid. 28. Ibid.: 23. 29. On the transcultural interactions along the Silk Road and other trans-Himalayan trade routes, see Strickmann 1982. For examples of northwestern Chinese monks heading for the western regions in early medieval China and the resultant exchanges, see Du 1995: 103113. Sharf acknowledges the routes of exchange connecting South and Central Asia to China but immediately dismisses their signicance because travel between these two regions was difcult if not impossible. Furthermore, he argues that foreign monks with mastery over Buddhist scripture and doctrine . . . were relatively few in number, and their command of Chinese was often wanting. And while some Chinese pilgrims did successfully

Problems and Perspectives

11

most part did not see themselves as so culturally isolated that they could not comprehend Buddhism as it was understood in India, even if they were reading only Chinese translations of Indic scriptures. In fact, early Chinese Buddhists were frequently dissatised with translations of Buddhist texts and sought to improve translation techniques.30 Although it is senseless to speak of an unchanging core of Buddhism perpetuated across cultural and geographical boundaries, it is equally problematic to negate the meaningful relationships and rich interchanges that existed among different Buddhist cultures and end up implying the existence of pockets of relatively isolated Buddhisms that were culturally severed from one another and virtually inaccessible to one another. My investigation begins directly with Dizangs appearance in fth-century China. However, the Indian and Central Asian evidence of Kitigarbha is discussed in the appendices so that meaningful connections (or their lack) can be considered in relation to the making of Dizang Bodhisattva. This structural organization is a strategy to circumvent the kind of master narrative Sharf critiques, which always opens the historical account of Chinese Buddhism with its Indian origins, followed by its subsequent domestication in China. On the other hand, modern scholarship has already witnessed the critical aw of sinological readings that do not sufciently attend to the preChinese Buddhist elements and are overly quick to seize on the indigenous cultural conditionings in Chinese Buddhism. A classic example is the case of Buddhist lial piety, touted for a time as a Chinese Buddhist innovation, until Indologists called attention to antecedent practices in Indian Buddhism and shattered the myth that Buddhist lial piety was unique to China.31 By presenting relevant Indian and Central Asian Buddhist examples in the appendices, I attempt to avoid the pitfalls of either studying Chinese Buddhism in a contextual void or overly emphasizing the role of earlier Buddhist cultures. Although the danger of indiscriminately ltering Chinese Buddhism through Indian lenses should be acknowledged, the encounter metaphor is really problematic only when it presumes a linear one-way evolution of Budjourney to India . . . only a handful are remembered in the historical record for their contributions to the transmission of Buddhism to China (2002a: 18). Hence he concludes, [T]he educated Chinese elite, not to mention the unlettered masses, remained largely ignorant of the vast linguistic and conceptual divide that separated them from the world of Indian Buddhism (ibid.: 21). 30. For a discussion in Chinese on issues in the translation of texts in early Buddhist China, see Tso 1992. Tso Sze-bong explores the extent and quality of Sanskrit language skills, the dynamics of translation teams, and the evolution of translation practices over time. For English discussions of translation activities in early Chinese Buddhism, see Boucher 1996, 1998; Zrcher 1977, 1991. 31. For classic treatments of lial piety as a patently Chinese Buddhist practice, see Chen 1968; 1973: 1464; Jan 1995. Schopen (1984) and Strong (1983) countered the assumption that Buddhist lial piety was largely a sinitic development by highlighting examples of lial behavior in Indian Buddhism. On the cult of lial piety in medieval China, see Knapp 2005. On lial piety in Chinese history, mythology, and thought, see the essays in Chan and Tan 2004; Ching and Guisso 1991.

12

Introduction

dhist history whereby Chinese Buddhist developments are always traced to some Indian original. As previous scholarship has pointed out, the medieval interaction between non-Chinese and Chinese Buddhist cultures involved movements in multiple directions that included retroactive inuxes of artifacts, ideas, practices, and texts traveling out from China along the Silk Road.32 One example is the Prajpramithdya-stra (Bore poluomiduo xin jing , Scripture on the Heart of the Perfection of Wisdom), which Jan Nattier has argued could have been composed in China and transmitted back along the Central Asian trade routes.33 Vestiges of Chinese Buddhism were present in Turfan during the ninth century, and cave art in Khara Khoto (Heishui cheng ), dating from the eleventh century onward, further documents sinitic presences.34 Any discussion of the Dizang cult should thus account for relevant elements in this fertile cross-cultural interchange, even while acknowledging that Dizangs mythology and individuality were largely dened in the cultural and social contexts of China. For example, Indian elaboration of Kitigarbha in the form of esoteric Buddhist iconography, practice, and texts made its way into China around the eighth century, after the Dizang cult had gained a rm foothold in the Chinese environment. Originating outside of China, these expressions of the bodhisattva evidently intermingled and melded with the extant art, history, mythology, ritual, and texts of the Dizang cult. Ultimately, these diverse religio-cultural
32. See Nattier 1990. Also see Du 1995: 142150, for the impact of Northern Liang Buddhism on religion in Kharakhoja (Gaochang ). Kharakhoja was the capital of the ighurs when they moved into the Xinjiang region from Mongolia in the ninth century. 33. Nattier (1992) points out that the Heart Scripture gained popularity in China well before it became the subject of commentarial attention in India and that it has maintained a central role in East Asian Buddhism from the seventh century to the present. Reconstructing its textual history, Nattier argues that the Heart Scripture may be a Chinese apocryphal text composed of an extract from the Large Stra of Kumrajva (itself a translation of the Indian Pacaviati-prajpramit-stra) together with an introduction and conclusion composed in China. According to Nattier, the scripture might have made its way back to the Sanskrit milieu through Buddhist pilgrims to the west. 34. The existence of snippets of a ighur manuscript that contain excerpts from the Dizang pusa benyuan jing suggests that this scripture made its way from China to Turfan; see Zieme 1990: 379384; facsimiles: plates III. Leaets from a Tangut manuscript of the scripture were also found recently in the northern section of the Mogao caves at Dunhuang. Maria Rudova (1993) argues that Chinese-style paintings from Khara Khoto in the State Hermitage Museum, especially those associated with the Amitbha and Guanyin cults, were executed under the inuence of Chinese art of the Song period, when relations between China and the Tangut empire were at their peak. Khara Khoto (Mongolian for Black City) was an important city of the Tanguts, a tribe of Tibetan origin whose homeland was originally in the highlands of western Sichuan. In the eleventh century the Tanguts took control of the Hexi corridor, the section of the Silk Road leading west from central China; by the 1070s, Tangut rule had extended to Dunhuang. Khara Khoto was excavated from the sands of what is now a largely uninhabited region of the Gobi by Russian expeditions led by P. K. Kozlov in 19081909 and 19231926. Among the city ruins was a large stpa; its base contained a treasure trove of Buddhist sculpture, scriptures, and painted silk banners, all of which are now in the collection of the State Hermitage Museum in St. Petersburg. For a study of Khara Khoto, see Piotrovsky 1993.

Problems and Perspectives

13

strands became so tightly interwoven into the complicated fabric of the Dizang cult that they can seldom be clearly differentiated into disparate cultural threads. Hence, when references in this study are made to Indian, Central Asian, or Chinese elements, they should always be understood to be embedded in the expressions of Chinese Buddhism.

Lord of the Underworld

Dizangs connection with the infernal realm is taken so much for granted in Chinese religion today that, as we saw, the ritual participants and observers at the Buddhist Lodge simply assumed that the ritualist was enacting Dizangs descent into hell. For many Chinese, Dizang Bodhisattva holds out the promise of divine intercession for the welfare of deceased relatives and for the pacication of the dead in general. At the same time, because of Dizangs intimate association with the realm of the dead, this bodhisattva is regarded with some ambivalence. In temples, his isolation in a separate cloister reserved for the tablets of the deceased, away from the main shrine hall, spatially symbolizes the curious stigma attached to him. Despite his evident popularity, he is rarely enshrined in household altars, and lay Buddhists are hesitant when it comes to reciting the Scripture on the Past Vows in their homes.35 It is feared his presence might invite into the household malignant forces causing illness, mishap, death, and ghost possessions.36
35. So dominant is the underworld Dizang in contemporary Taiwan that Buddhist leaders have taken measures to correct what they deem to be misconceptions about him. In an effort to reinstate his proper status as a bodhisattva, Buddhist modernist intellectuals in Taiwan have reinterpreted Dizang in terms of the this-worldly ideals of so-called Humanistic Buddhism (renjian fojiao ). A prominent example is Yinshun (19062005), one of the foremost proponents of Humanistic Buddhism, who composed an essay titled Dizang pusa zhi shengde ji qi famen . His writing seeks to dispel widespread misconceptions about Dizang and reasserts the bodhisattvas vows and altruism; see Yinshun 1989. Another example is the nun Zhengyan (1937), the founder of the Buddhist Compassion Relief Meritorious Association (Fojiao Ciji gongdehui ), a transnational Buddhist charity organization. Under her leadership, an image of Dizang is placed together with those of kyamuni Buddha and Guanyin in the main shrine at Cijis monastery, the Abode of Still Thoughts (Jingsi tang ), located in Hualian , eastern Taiwan. In Zhengyans interpretation, Dizang represents the great vows of the bodhisattva mentioned in the Scripture on the Past Vows and is intimately related to this world; see Zhengyan 2003. Elsewhere I have argued that the appearance of the so-called Sah Triad (suopo sansheng ), consisting of kyamuni anked by Guanyin and Dizang, in contemporary Taiwan, should be understood as furnishing appropriate visual imagery and devotional practice to embody the this-worldliness of Buddhist saints; see Zhiru 2000. 36. On the phenomenon of avenging ghosts in Chinese religion and society, see Feuchtwang 2003; Nickerson 1997; Poo 1998: 4168. Vengeful spirits (yuanhun ), that is, the spirits of those unjustly killed who return to seek vengeance on their killers, is a pre-Buddhist Chinese concept. With the coming of Buddhism, this indigenous belief in vengeful spirits was combined with the doctrine of karma and retributive rebirth. Stories dealing with avenging spirits appear early in the Soushen ji composed in 335345. For a study of the Soushen ji, see Dewoskin and Crump 1996. An important compilation of stories on aveng-

14

Introduction

Uncritical acceptance of this underworld imagery on the part of modern scholarship is partially responsible for the curious paucity of research on Dizang in China. Except for Marinus Willem de Vissers now-dated survey (1914), no book-length study was undertaken until the last decade.37 Dizang has received minimal attention from Buddhologists, even though he is an exemplar of the bodhisattva ideal, a core Buddhist doctrine. His association with the realm of the dead places Dizang, more so than the other major bodhisattvas, in an uncomfortable fuzzy zone that crosses the threshold of grassroots Chinese religion, tying him inseparably to the afterlife and the pacication of restless spirits and endish demons.38 Consequently, Buddhologists tend to relegate Dizang to the category of folk or popular religion, where Buddhism has lost its distinctive contours and exists more or less as amorphous phenomena frequently subsumed under indigenous religions.39 They view the Dizang cult as a marginal topica stepchild of Buddhologyirrelevant to the study of orthodox or mainstream Buddhism and only pertinent to studying diffused religious culture in China. This attitude harks back to the distinction between so-called great and little traditions rst proposed in 1956 by the anthropologist Robert Redeld. The great tradition is considered a construct of the elite and literate, who consciously transmitted it to their successors, whereas the little tradition is perpetuated unwittingly by the largely illiterate peasants in traditional society. This two-tier paradigm is associated with a bifurcation of culture and society and enforces a polarization of the religion of the elite versus that of the popu-

ing ghosts, the Yuanhun ji , is attributed to the sixth-century writer Yan Zhitui (531591). For an English translation of these stories, see Cohen 1982. 37. I will return to survey trends in modern scholarship on Dizang Bodhisattva later in this chapter. In China, Dizang is the most popular Buddhist deity next to Guanyin, but he has attracted relatively few studies compared to the voluminous scholarship on the Guanyin cult. In western scholarship alone, several studies on Guanyin exist: for example, the early overview by de Mallman (1948); a close investigation of the Miao-shan legend by Dudbridge (1978); a cross-cultural study of the thousand-armed Guanyin by Chandra (1988); and the previously mentioned work by Y (2001). Guanyin is also an exceedingly popular subject in art history; for a recent study in English, see Karetzky 2004; for a study of the water-moon Guanyin, see the unpublished dissertation by Chan (1996) and the unpublished paper by the German art historian Petra Rsch (2006). Also see Guoli gugong bowuyuan 2000 for the catalog of a special exhibition of Guanyin images held at the National Museum of Taiwan in Taipei. Several important monograph articles on Guanyin have been published: for example, Stein (1986) investigated the emergence of Guanyin as a female deity; Angela Howard (1990) studied Guanyin iconography during the Tang and Song periods; Robert Campany (1996b) translated early miracle tales on Guanyin. There also exists an overwhelming body of research on Guanyin in Chinese and Japanese that cannot be listed here. 38. To a greater or lesser extent, the major bodhisattvas perform functions connected with death and the afterlife. For example, Guanyin appears frequently as a savior in the realm of death and the afterlife. In the Xiangshan baojuan , dated to the sixteenth century, Guanyin is led on a tour of hells. See Dudbridge 1978; Y 2001: 320333; Overmyer 1999: 3447. 39. For discussions of folk religion, especially in relation to Buddhism, see Overmyer 1990; Teiser 1995, 2004. On popular religion in China, see Bell 1989; Feuchtwang 2001.

Problems and Perspectives

15

lace.40 Scholars have repeatedly agged this problematic bifurcation and the stereotypical characterization of religion and culture it engenders, but the rippling effects continue to be felt in the study of religion. One example is the tacit assumption that the Dizang cult is some form of religion for the masses and not quite on the same footing as elite Buddhism. This perspective overlooks the crucial fact that Dizang is a bodhisattva who has retained throughout his career obvious orthodox doctrinal overtones. Dizang thus dees any attempt to segregate elite doctrine from cultic practice; rather, he exemplies and demonstrates continuity not only between doctrinal and devotional elements, but also between so-called elite and folk religion. Although it is true that Dizang is the focus of cultic practice, it is also the case that he draws power from his association with core doctrinal concepts of elite Buddhism. To label him univocally as either elite or folk cannot convey fully the complex dynamic in the making of the Dizang cultor for that matter, the making of any Buddhist cult. For example, as discussed in Chapter 5 of this study, Dizangs underworld character may have initially appeared in narratives associated with Yogcra and Avatasaka (Huayan ) teachings, in which the three realms of existence, including hells, are created by the mind alone. Representing Dizang solely as an underworld sovereign reduces the complex dynamic of his cults history to a more or less linear development of his underworld aspect. From this perspective, the rise of Dizang worship is interpreted as a folklorization of the bodhisattva ideal. But the image of Dizang as savior of the damned hardly exhausts his many functions in Chinese religion and society. In medieval religion, he was a multivalent gure situated at the nexus of religious experimentation, and the medieval cult of Dizang embodied a broad spectrum of elements extrapolated from diverse religious sources.41 Even in the modern period, Dizang, as savior of the damned, cannot account for the range of practices associated with his cult. The tremendous output of Buddhist literature in Taiwan today includes a series of liturgical manuals known as the Fojing xiuchi fa (Method of Upholding
40. Note that Redeld himself cautiously qualied his paradigm with an observation that in many traditional societies, China included, there existed overlaps and mutual exchanges between the two traditions. For application of this two-tier paradigm to Chinese religion, see Sangren 1984 and Jochim 1988. Applying the paradigm to Theravada Buddhism, Spiro (1970) distinguishes between kammatic Buddhism and nibbanic Buddhism, which are then correlated to folk and elite expressions. 41. In fact, several short studies have already called attention to salient aspects of Dizang Bodhisattva apart from his underworld function: Yabuki (1927) argued for an early connection with the medieval Buddhist school Sanjie jiao; as part of his investigation of esoteric Buddhism, Osabe (1982) studied a hybrid Buddho-Daoist exorcism formula attributed to Dizang; Powell (1987, 1993) investigated the pilgrimage cult of the Golden Dizang on Mount Jiuhua; Lai (1990) and Kuo (1994b) analyzed the Zhancha jing to illustrate Buddhist integration of indigenous Chinese practices; most recently, Zhiru (20012002, 2005) traced the medieval associations of Dizang worship with the cults of Amitbha and Maitreya.

16

Introduction

Buddhist Scriptures), formulated for daily Buddhist practice. Volume 12 in this series focuses on the Bodhisattva Dizang.42 The prescribed daily practice consists of a liturgical sequence commonly found in Chinese Buddhist rituals plus a two-step meditation procedure of calm and insight (Ch. zhiguan ; Skt. amatha-vipayan) featuring Dizang as the object of meditation.43 In addition, contemporary Buddhist preachers in China continue to deliver sermons on the Zhancha shane yebao jing (Scripture on Divining the Retribution of Skillful and Negative Actions), a sixth-century text on Dizang worship, popular divination, and karma composed prior to the emergence of his underworld aspect.44 Esoteric echoes of Dizang also reverberate in religious milieux where Chinese and Tibetan Buddhism are increasingly interfused; Tibetan monks in exile, for example, preside over Dizang repentance ceremonies when addressing the needs of Taiwanese society and immigrant Chinese communities in Southeast Asia. The savior of the damned image thus obscures the multivalent character of Dizang worship.

The Study of Bodhisattva Cults

Although monographs have occasionally appearedfor instance, Marinus Willem de Vissers studies of Dizang (1914) and kagarbha (1931); Raoul Birnbaums study of Majur (1983); and Alan Sponberg and Helen Hardacres edited volume on Maitreya (1988)the investigation of bodhisattva cults and their signicance in Buddhist history has been curiously marginalized in Buddhist studies and remains largely the province of art historians.45 In recent decades, however, an exciting shift in the study of Chinese Buddhism has brought about a welcome surge of scholarly interest in Buddhist cults. Endeavoring to apply cross-disciplinary approaches, Buddhologists have increasingly engaged in dialogue with art historians and undertaken serious investigation of Buddhist art and epigraphy in their research and writing. In the subeld of Chinese Buddhism, this trend is further expedited by the growing opportunities for scholars outside of China to conduct eld investigations at Chinese archaeological and cave sites. One heartening result
42. The volume is titled Ruhe xiuchi Dizang jing ; see the listing of primary sources in Works Cited. This manual invokes the Scripture on the Past Vows of Dizang Bodhisattvanot for its connection with the cult of the dead, but for its presentation of the bodhisattva ideal and the practice of lial piety. 43. Ruhe xiuchi Dizang jing, 5660, 6771. 44. Zhancha shane yebao jing, T839:17.901c910c, ascribed to Putideng (Bodhidpa?). I will return to discuss this scripture in Chapter 3. A contemporary Chinese preacher, widely known as the monk elder Mengcan , delivered a sermon on the text, which has been recorded, printed, and circulated under the title Zhancha shane yebao jing jiangji . One of his lectures is also printed in the form of a tract manual titled Dizang zhancha lun xiuxing shouce , which comes with a packet of nineteen hexagon-shaped wheels to be used for Dizang divination. 45. Buddhist art history produces the bulk of the studies on buddha and bodhisattva cults. For early studies of Avalokitevara and Majur, see de Mallmann 1948, 1964.

Problems and Perspectives

17

is a owering of scholarship on the bodhisattva cults, many of whose histories are best documented using visual and material evidence. Y Chn-fangs tome on Guanyin, Franoise Wang-Toutains monograph on the Bodhisattva Dizang, and Robert Gimellos research on Majur and Cund (Zhunti ) all include investigations of visual materials.46 The signicance of studying bodhisattva cults is perhaps best understood in light of the reasons for its relative neglect in past scholarship. One reason is the previously mentioned tendency to regard religious history as the progression of ideas and texts.47 From this perspective, the bodhisattva is viewed largely as a doctrinal expression and the focus is on mapping typologies for its changing explications in scriptural writings and philosophical treatises.48 This regnant tendency in older Buddhist studies scholarship also meant that the bodhisattva category attracting the most scholarly attention was the jtaka bodhisattva (bensheng pusa ) and the stages on the path entailed in its 49 conceptualization. The superhuman savior bodhisattva was treated only as the culmination of bodhisattva practice. Moreover, textbook accounts too often bifurcate the presentation of the bodhisattva ideal into doctrinal statements of the bodhisattva path and devotional practices pertaining to the cults of savior bodhisattvas, mistakenly engendering the impression that doctrine and devotion are mutually exclusive aspects of religion.50 In fact, the

46. Y 2001; Wang-Toutain 1998; Gimello 1997, 2004. 47. On the primacy of text in the study of culture and religion, see Sullivan 1990: 4159. The necessity to integrate the study of Buddhism in China more fully within Chinese studies as a whole has been noted by several Buddhologists; see Buswell 1990: 13; Gregory 1999: 1920. 48. One typology of the bodhisattva doctrine is Lewis Lancasters fourfold stratication, enumerated as follows: (1) the jtaka bodhisattva, or the aspiring bodhisattva on the arduous path to buddhahood; (2) the phantasma bodhisattva, represented as a savior gure larger than life; (3) the audience bodhisattva, who graces the buddha assemblies; and (4) the living bodhisattva, who appears as an incarnation in this world. See Lancaster 1981: 153 161. To some extent, the four categories often overlap. In early medieval China, Dizang seems to have appeared rst as an audience bodhisattva, who then red the religious imagination of Chinese Buddhists so much so that he was rapidly transformed into a focus of worship, a savior bodhisattva. The Scripture on the Past Vows contains ashbacks into the past lives of this bodhisattvathat is, his career as a jtaka bodhisattva. Finally, the cult of the Golden Dizang at Mount Jiuhua is centered on his manifestation as an incarnated, or living, bodhisattva, the Korean ascetic Kim Chijang. 49. For a classic example, see Jaini 1988. Drawing from different texts, Jaini pieces together a cogent presentation of Maitreyas bodhisattva path. For a survey of textual presentations on the bodhisattva doctrine, see Dayal 1970. The same can be said for most of the papers in the volume on the bodhisattva doctrine edited by Kawamura (1981). 50. The classic example is Paul Williamss (1989) excellent survey of Mahyna Buddhism. In Chapter 9, titled The Path of the Bodhisattva, Williams highlights the doctrinal aspects of the bodhisattva ideal, whereas Chapter 10 is titled Faith and Devotion: The Cults of Buddhas and Bodhisattvas. Another textbook survey of Buddhism (Harvey 1990) dedicates seven chapters to Buddhist doctrine and four chapters to Buddhist practice. The cults of Avalokitevara, Amitbha, and Bhaiajyaguru are treated in a chapter titled Buddhist Practice: Devotion. In an essay, F. V. Tiso (1993) discusses the emergence of the bodhisattva in terms of the Mahynist systematized presentation of the bodhisattva path.

18

Introduction

very phenomenon of the savior bodhisattva already implies a relationship between bodhisattva doctrine and cultic practice.51 In part, the marginalization of religious devotion can be traced to Protestant rhetoric that condemns Catholic excesses meant to satiate the ritualistic and emotional needs of the uncritical masses in presumed opposition to the basic spirituality of Protestant Christianity.52 This disdain of devotional practice and the bifurcation of doctrine and devotion are hardly true of medieval Christianity, much less of Buddhist cultures, and western scholars have increasingly called attention to the Protestant biases in western scholarship on Buddhism.53 David Snellgrove, for instance, has identied the problem of exaggerating the historical Buddhas humanity, cautioning that the Buddha is always more than human to his followers in Asia, who do not view the humanity and transcendence of the Buddha as mutually incompatible.54 In other words, the Buddha does not t the polarized paradigm of human and divine that underlies Christian discourse. Accordingly, scholars must exercise caution in applying to the Buddha those religious categories styled after western models of religion. I would argue that the same should be said for the savior bodhisattvas: The tendency to separate the doctrine of the bodhisattva ideal from the devotional cult of the savior bodhisattva is misleading. Conning these cults to the overarching category of devotionalism as understood in modern western discourse fails to capture the complex array of religious attitudes associated with this Buddhist phenomenon or its historical signicance. Interpreting the deication of the bodhisattva largely as a strategy to broaden the appeal of the bodhisattva ideal reduces the concept of the savior bodhisattva to a secondary expression.55 Instead, the concept of the savior bodhisattva should be under51. Bielefeldt (2005) surveys the term practice (including cultic practice) as understood by the Buddhist traditions and Buddhist studies; his observations on the nuanced (and often embedded) relationships between practice on the one hand and theory (principle) or ideal on the other are also applicable to the problem of separating the bodhisattva doctrine from the cult of the savior bodhisattva. 52. Art historians studying American Protestant art have shown that despite Protestant criticisms of icons and devotionalism, the use of religious art persisted in church and domestic settings in conjunction with worship; see Morgan 1996, 1998, 2001; Morgan and Promey 2001. Also see Finney 1999 for essays on the role of visual arts in the Calvinist tradition. 53. Anthropologists rst called attention to the Protestant (and Orientalist) biases in the study of Buddhism; see Almond 1988; Gombrich and Obeyesekere 1988: 202240. Subsequently, scholars in Buddhist studies have discussed Protestant implications in the study of Buddhism in the west: see Schopen 1991a; Lopez 1995; Strong 2004: 15. 54. Snellgrove 1987: 2935. On the representation of the Buddha, particularly in the west, see Lopez 2005. 55. The demythologization of the savior buddha and bodhisattva is certainly present in Buddhist discourse, particularly in modernist Buddhist discourse that emphasizes the goal of buddhahood as a humanistic endeavor, but also in Chan Buddhist rhetoric. However, as Faure (1991, 1992) has shown, Chan Buddhists were associated with cultic practices in the mummication and veneration of awakened Chan teachers. On the ritual aspects of Chan Buddhism, see Foulk 1993; Foulk and Sharf 19931994. On the role of Confucian ritual and

Problems and Perspectives

19

stood as richly complex, drawing its power precisely from juxtaposing multiple strains of religious expression (art, doctrine, institution, myth, praxis, ritual) and thus rendering the bodhisattva ideal accessible to an audience of diverse backgrounds and applicable to a variety of social contexts. Rather than a degeneration of Buddhist doctrine into some kind of secondary religion for the gratication of the masses, the cult of the savior bodhisattva transforms the abstraction of the bodhisattva doctrine into a concrete, everyday realitya living presence that believers can tangibly experience through iconography, myth, ritual, and social institutions.56 Doctrine and devotion, transcendence and humanity, are inextricably entwined to create a potent religious expression that cannot be reduced to any single dimension of religion. Another reason for past scholarly neglect of bodhisattva cults is also a component of the master narrative of the history of Chinese Buddhism to which Sharf calls attention, namely, undue xation on lineages, schools, and sects and the persistent representation of Chinese Buddhist history in terms of their vicissitudes.57 Sharf argues, rst of all, that the boundaries were never as clear as once thought; and second, that in several cases the lineages and schools do not correlate to actual social institutions but are products of sectarian polemics and loose retrospective classications, many of which derive from the sectarian structure of Japanese Buddhism. More often than not, bodhisattva cults are looked upon as marginal forms that evolved on the periphery of doctrinal explication, relevant only insofar as they render the teachings of a lineage or school more accessible to a larger audience. Moreover, cultic expressions often undermine the reied demarcations of lineages and schools. In fact, precisely because bodhisattva cults are prone to blur such proposed boundaries, the study of these cults should have corrective ramications for the old understanding of Chinese Buddhist history. Cults better document the complex dialectics and interrelationships underscoring East Asian Buddhism than do the loose typologies traditionally deployed. For instance, as the present study shows, the medieval cult of Dizang presented an array of relationships with diverse forms of Buddhism at various junctures in its history. In the eighth and ninth centuries, the Dizang cult appears to have possessed simultaneous alliances with Teaching of Three Levels (Sanjie jiao ) and Pure Land (Jingtu ), although polemical literature pits these two schools as rivals with clearly demarcated objects of worship and practices.58 Buddhist cults thus provide material evidence for Sharfs argument that lineages, schools, and traditions did not exist as discrete entities. However, one must be careful not to overstate the case and deconstruct the lineages in China altogether. They did exist insofar as they represented accepted doctrinal or soteriological distinctions for particular
ancestor worship in Chan Buddhism, see Jorgensen 1987. For the veneration of images in modernist expressions of Taiwanese Buddhism, see Zhiru 2000. 56. See Campany 1993b. 57. Sharf 2002a: 78. 58. See Zhiru 20012002: 8085.

20

Introduction

groups of Buddhists in China (especially after the Song period), even if they did not always correlate with institutional or social realities.

The Study of Dizang Bodhisattva

Until the late 1990s, Japanese scholars produced the bulk of scholarship on Dizang Bodhisattva. In his manifestation as Jiz , Dizang pervades Japanese society to such an extent that his fame surpasses that of Guanyin, known as Kannon in Japan, who is usually the more popular of the two in East Asia. From the late Heian period (8981185), Jiz worship increasingly permeated Japanese culture and society, generating a broad range of Jiz manifestations endowed with specic names and sometimes peculiar iconographies. Hence, the religious landscape of Japan is dotted with shrines and temples dedicated to Anzan Jiz (Easy Birth Jiz), Hara-obi Jiz (Belly-girdle Jiz), Enmei Jiz (Life Prolonging Jiz), Ind Jiz (Jiz Who Guides [the Deceased to Their Afterlife Destinations]), and Roku Jiz (Six Jizs), to name just a few.59 Although Dizang iconography in medieval China was wide-rangingsometimes he appears as a monk, a sovereign, or even a royal householdermedieval Japanese exhibited an early preference for Jiz the monk, who over time became Jiz the novice. In a sense, Jiz in Japan was deliberately downscaled from the remote, otherworldly bodhisattva to the friendly, approachable little monk whom one could expect to run into on the next street corner.60 Given the preeminence of Jiz in Japanese culture, religion, and society, it is not surprising that Japanese scholars should be fascinated with this bodhisattva and should have produced a wealth of excellent research on him.61 This large corpus of scholarship, however, is interested in the Chinese development only insofar as it relates to the origins of the Jiz cult. Given that the Japanese Jiz and the Chinese Dizang are in many respects quite different, this tendency frequently alters the Japanese presentation of the Chinese historical reality. An important exception is the pioneering scholarship Japanese have produced on Dunhuang art that includes studies of Dizang iconography, notably the seminal contributions of Matsumoto Eiichi (1933, 1937). Although there has been a lack of scholarship on Dizang Bodhisattva in China, the past decade has witnessed a welcome shift in scholarly interest that reects a heightened awareness of the signicance of this bodhisattva in Chinese religion. Besides a multiplying collection of articles, mostly on Dizang art in China, four book-length publications have appeared: (1) Franoise WangToutains Le Bodhisattva Kitigarbha en Chine du Ve au XIIe sicle; (2) Pan Liangwens Zhongguo Dizang pusa xiang chutan ; (3) Zhuang Mingxings Zhongguo zhonggu de Dizang xinyang
59. For studies on Jiz in Japan, see Fuji 1974; Hayami 1975; Ishikawa 1995; Tanaka 1989. 60. Laeur 1992: 47. 61. See Manabe 1960; Sakurai 1983; Mochizuki 1989; Ishikawa 1995.

Problems and Perspectives

21

; and (4) Zhang Zongs Dizang xinyang yanjiu .62 Although these studies on Dizang were all published within a ve-year period, they were apparently conducted independently and in relative isolationwith the exception of Zhangs study, which does cite the other Chinese works. Trained as an art historian, Pan Liangwen has collected the visual evidence and produced a slim volume surveying Dizang art and inscriptions in China. Zhuang Mingxings publication, based on his M.A. thesis, discusses a range of materials that are often treated more substantively in the other publications. Zhang Zong, a mainland Chinese scholar, exhibits a superb mastery of the archaeological and art-historical materials, and he provides an exhaustive survey of Dizang sculpture, paintings, and cave art in China and Central Asia. His contribution is especially important for its inclusion of recent archaeological ndings that have not been published outside of China. In addition, his survey of texts includes a section in which he discusses the Dizang baojuan (Precious Scrolls on Dizang), a relatively unstudied collection of literature especially crucial for exploring Dizang in late imperial China.63 Of the studies on Dizang, Zhangs work is the most ambitious in scopenot only in terms of geography and chronology, but also in the staggering wealth of literary and visual sources he has assembled.64 The French scholar Franoise Wang-Toutain traces the development of Dizang Bodhisattva up to the twelfth century. Possessing an excellent command of Buddhist art, philology, and history, she analyzes a substantial corpus of canonical and noncanonical literature, the bulk of which is scriptural. A major contribution is her extensive cross-comparison of Dunhuang manuscripts, which sheds light on the complicated and often ambiguous histories of Dizang texts. Her cataloging of Dunhuang manuscript copies of scriptures on Dizang will remain an immensely valuable resource for all subsequent studies. Wang-Toutains monograph is divided into four sections: texts (scriptures), personality, cult, and iconography. The chapter on Dizang iconography is based largely on Dunhuang art, although examples from other regions are included where relevant. For Wang-Toutain, Dizang iconography is chiey used as collaborative evidence for the conclusions she has already
62. Art historians, mostly from China but also from Taiwan, have increasingly studied Dizang art, particularly from Chinese Buddhist cave sites. Meriting special mention is an essay by Paul Katz (2003), who offers an anthropological study of the growth of the Dizang temple (Dizang an ) in Taiwan during the period of Japanese occupation. His study is largely a statistical investigation that focuses on economic and political conditions shaping the local development of the Dizang temple and does not treat Dizang Bodhisattva or the religious practices in any signicant manner. But its focus on the modern period takes Dizang study in a new direction. 63. The literary genre baojuan, or precious scrolls, is relatively less studied in western scholarship. For an important volume on the precious scrolls in the sixteenth and seventeenth centuries, see Overmyer 1999. 64. Although his principal focus is China, Zhang also surveys historical evidence from Central Asia, Korea, Japan, Vietnam, and Tibet. His work thus serves as an invaluable guide to the resources for studying Dizang Bodhisattva.

22

Introduction

reached in her textual analysis.65 Although the monograph is not organized chronologically, her solid investigation does map a more or less linear trajectory whereby venerating Dizang originates in the cult of eight bodhisattvas in Central Asia and culminates in its Chinese underworld expression. One can thus isolate three phases in her mapping of Dizangs history. The rst phase is a connection with the concept of the decline of Buddhism, apparent in the Scripture on the Ten Wheels, which she traces to Central Asia through scriptures ascribed to the translator Narendrayaa (Naliantiyeshe , 517589). The second phase is the dissemination of Dizang worship in the seventh century due to Sanjie jiaos promotion of the Scripture on the Ten Wheels and the teaching of the Final Dharma. The third phase is the ascendancy of Dizang worship after Sanjie jiaos disappearance from the religious scene in the ninth century, at which time the Dizang cult increasingly blended with other cults, notably those of Amitbha and Guanyin. Within this context, Dizang steadily emerged as the Buddhist savior of the damned, superceding claims to this role by other bodhisattvas, including the popular Guanyin Bodhisattva. It is telling that the chapters in which Wang-Toutain discusses Dizangs personality and cult focus largely on his underworld role. Although I hesitate to oversimplify what is an exceedingly rich investigation of complex developments surrounding the three phases of Dizangs history, Wang-Toutains study clearly assumes that the development of Dizang was a steady folklorization resulting in his role as savior of the damned. Why this sudden scholarly interest in the Dizang cult? Given the considerably smaller corpus of writings on Dizang relative to those on other major bodhisattvas, archaeological and art-historical ndings have been especially instrumental in illuminating neglected aspects of his cult and expanding our understanding of Dizangs history. Although the potential of Dunhuang art and texts in this regard has been acknowledged for some time, the wealth of evidence found at other Buddhist sites has been relatively unstudied until recently. The art at these sites turns out to reveal fascinating and rarely recorded manifestations of Dizang worship. For instance, I have shown elsewhere that art and epigraphy at Sichuan and Dragon Gate (Longmen ) indicate Dizangs link to the Amitbha Pure Land, while Shaanxi iconography hints at a plausible connection to the Maitreya cult.66 As an object of study, Dizang worship constitutes a window for glimpsing the interrelationships between divergent religious expressions, otherwise mistakenly regarded as discrete and separate strands. Hence, the study of Dizang potentially has larger implications for the study of Chinese Buddhist history. Because modern scholarship analyzes Dizang Bodhisattva largely from the single perspective of his popularized role as savior of the damned, it achieves only a partial reconstruction of the medieval Dizang cult.67 The need for a study
65. Wang-Toutain 1998: 259. 66. Zhiru 20002001; 2005. 67. Wang-Toutain introduces her study as an investigation of the Bodhisattva Dizang as the Teacher of the Desolate and Dark [Region]or, in other words, the savior of the damned (1998: 57).

Problems and Perspectives

23

of Dizang in medieval China that explicitly addresses the implications of the richly varied expressions of Dizang worship and so reconsiders its role in Chinese Buddhism is thus imperative. The Dizang cult intersected not just with different forms of indigenous religion (religious Daoism, Confucianism, Chinese divinatory and shamanic practices), but also with diversied strands of Chinese Buddhism (esoteric Buddhism, Maitreya mythology, Pure Land, Sanjie jiao). The process of cross-cultural assimilation that the Dizang cult embodies was frequently punctuated by tension and conict. Against this complicated religious background, the image of Dizang as Lord of the Underworld came to the forefront in the late eighth or early ninth century. Yet this aspect never completely subsumed the others, which persisted into the tenth century and after.

Methods and Sources

Whereas historical documentation is scarce in India, scholars of Chinese Buddhism are fortunate to have at their disposal an abundance of historical records, ranging from catalogues to Buddhist historical writings. Risking an overgeneralization, one may say a preoccupation with historical documentation has always been an outstanding feature of Chinese culture. Nonetheless, the substantive corpus of literary materials is a mixed blessing, for it quite frequently conceals as much as it reveals.68 The majority of surviving written records in the various archives of the Chinese Buddhist canon were composed by elite clerics, who dened normative or mainstream Buddhism and oversaw its transmission. These ofcial historical records should thus be perused with caution, and the task of the modern historian is much like that of an archaeologist, excavating the embedded layers of discourse with an eye to extricating silenced or misrepresented elements. The scholar must endeavor to explore variant expressions that unfolded outside the orthodoxy articulated in the transmitted archives of canonical works.69 In this regard, the recent owering of activities and studies based on archaeological and art-historical material is a boon to the study of Chinese Buddhist history. The recovery of Dunhuang archival manuscripts and art has already taught us that what we know about Chinese religious history is precious little in the face of what we do not know. It is not an overstatement to say that scholarly understanding of Chinese Buddhist history has drastically changed since the landmark discovery of the hidden archival chamber

68. As Erik Zrcher put it, Our view of Chinese Buddhism as a historical phenomenon is greatly obscured by the abundance of our source materials (1982b: 161). 69. Erik Zrcher also highlighted this point when he cautioned that the Buddhist Canon is the nal product of many centuries of clerical censorship after a constant process of expurgation (or even wholesale destruction) of . . . heretical texts (1982a: 168). As a corrective, Zrcher emphasized the need to investigate Chinese Buddhism in relation to indigenous religion particularly through the study of what he calls Chinese apocryphal literature. For a discussion on the polemical intentions that underscored and critically shaped the articulation of normative stances in Chinese Buddhist history, see Sharf 2002a: 1213.

24

Introduction

in Dunhuangs Mogao grotto 16 alone. For instance, our views of Sanjie jiao would still be conned to the glimpses available through polemical Buddhist literature if the Dunhuang manuscripts had not yielded a stack of Sanjie jiao writings. A recently discovered cave in Shaanxi where Sanjie jiao texts are inscribed on the walls promises to shed crucial light on the practices of that movement.70 Although we may not chance on another cache of manuscripts, scholars can look to the steady stream of reports on archaeological and art-historical ndings from elsewhere in China that are regularly published by the regional academies. The numerous Buddhist sites that house enormous collections of iconography and epigraphy furnish a wealth of data that can expand, or even revise, our perception of Chinese Buddhist history. As material embodiments of religion, archaeological and visual artifacts encapsulate actual religious practices in ways that written texts rarely do and frequently elucidate patterns of religion apart from the orthodoxy preserved in the received archives of transmitted writings. Accordingly, the principal thrust of my study of Dizang is to use extracanonical and visual materials to recover relatively neglected aspects of the cults history. This reconstruction is based on a close analysis of three categories of evidence: 1. Scriptures are undoubtedly a staple item in the study of religion. In my study, however, I have broadened the scope of scripture to include extracanonical as well as canonical texts. Especially central to this project is a genre of para- or noncanonical texts referred to as indigenous scriptures. The majority of texts about Dizang Bodhisattva, according to scholarly consensus, were composed in China. Particularly important are those scriptures recovered in the twentieth century from Chinese archaeological sites and from the archives of Japanese monasteries. Fortunately, some of these manuscripts are now published and accessible in modern editions.71 One key text is the Dizang pusa jing (Scripture on the Bodhisattva Dizang). Another important but less studied scripture is the Dizang dadao xin quce fa (The Exorcism Method of Dizangs Aspiration Toward Great Awakening).72 As mentioned previously, Buddhologists have only recently paid attention to art and epigraphy. An important source of Buddhist art is

2.

70. The cave is located at Jinchuan wan in Chunhua , Shaanxi, and is discussed in Chapter 2. For a preliminary investigation of the cave and its contents, see Zhang and Wang 2003. 71. Many of the texts recovered from Japanese monastery archives have been printed in modern editions of the Chinese Buddhist canon: the Dainihon zokuzky and the Taish shinsh daizky . 72. The Dizang dadao xin quce fa has often escaped the attention of Dizang specialists because the text uses deviant characters for the name of Dizang invented during Empress Wus rule (r. 684704).

Problems and Perspectives

25

3.

Dunhuang, which has yielded a sizeable collection of art pertaining to the underworld Dizang. For a more comprehensive understanding of the forms of Dizang worship, however, other art-historical sites must also be studied. For example, Longmen is important as a site housing the earliest Dizang art, while Shaanxi has yielded evidence of early visualizations of Dizang in connection to the Scripture on the Ten Wheels and possibly the medieval Buddhist movement Sanjie jiao. The art and epigraphy at Sichuan, a region that has attracted considerable attention from art historians in recent decades, conveys the broadest range of Dizang worship, some examples of which are exceptionally signicant because they embody religious strains documented in neither Dunhuang art nor textual sources. The genre of popular narratives known as miracle tales, both Buddhist and non-Buddhist, allows us to reconstruct medieval attitudes toward Dizang in the larger religious milieu. In this regard, the nonBuddhist miracle tales are especially valuable because they reect both a religious sensibility common across different segments of society and a diverse array of religious elements associated with Dizang worship. Especially important among the Buddhist collections of narratives is the canonical tenth-century compilation of Dizang miracle tales titled Dizang pusa xiang lingyan ji (A Record of Numinous Verications of Images of Dizang Bodhisattva; henceforth, A Record of Numinous Verications), completed in 989 by a member of the clerical elite.73 Compilations of miracle tales play a unique role in the religious canon by functioning as sub-texts articulatingalbeit in an edited and domesticated formthe beliefs and practices of the larger populace as they existed alongside the dominant traditions of elite orthodoxy.

Although these three genres of evidence are by no means exhaustive, they do constitute the major sources for the study of Dizang Bodhisattva in medieval China. Other sources of documentation, such as liturgical works and gazetteer records, are also mentioned when relevant. Although the present study adopts a historical framework as the main organizational principle, my approach is essentially interdisciplinary and explores the topic from as many perspectives as the scope of examined evidence allows.

Organization and Terminology

This study is divided into two parts and a concluding chapter. Part 1 introduces and analyzes the early manifestations of Dizang in China from the sixth to eighth centuries based on the early scriptural representations of Dizang in the Scripture on the Ten Wheels and the Xumizang fen (Section
73. The English translation of the title is derived from Teiser 1994: 43.

26

Introduction

on the Sumeru Treasury). The Scripture on the Ten Wheels inspired the basis for early cultic beginnings, including links to Sanjie jiao. In this early phase of the Dizang cult, the Chinese religious imagination was already at work crafting the identity and history of the bodhisattva, particularly in response to the socio-political conditions of fth- and sixth-century China. The Scripture on the Ten Wheels inspired innovative visual imagery, including the Shaanxi representation of Dizang as a buddha surrounded by the six paths of rebirth. Part 2 explores new directions in Dizang worship from the sixth to the tenth centuries, arguing that the array of Dizang images places him at the interstices of multiple religious trends during that period. Three categories of sources (scriptural writings, art and epigraphy, narrative literature) are investigated. Scrutiny of the evidence reveals that Dizangs history was hardly a straightforward evolution of his underworld function. Although this function was growing more visible, it remained one of severala testament to the vibrant diversity and rich complexity of Dizang cults. The Conclusion critically reexamines Dizangs role as Lord of the Underworld and afrms that Dizang was connected to several afterlife cultsnot just that of the underworld. In addition, Dizang formed important alliances with other religious expressions that did not fall under the rubric of afterlife practice, which suggests that Dizang may have had a larger role in the medieval history of Chinese Buddhism. Before embarking on the study proper, a few clarications concerning terms and concepts used in this study are in order. First among these is the potentially problematic designation medieval China. The phrase is employed simply as a term of convenience to avoid cumbersome repetitions of centuries or dynasties and should not be taken too literally.74 It does not intend any stance concerning the thorny issue of when the medieval period ended and the premodern, early modern, or late imperial periods began in China. Also, as may by now be clear, Dizang and Kitigarbha are alternative names for the same bodhisattva: The former name is used in the Chinese context; its Sanskrit counterpart is used with reference to Indian and Central Asian contexts. The chronological overlap of Part 1 (from the sixth to eighth centuries) and Part 2 (from the sixth to tenth centuries) also requires some explanation. Historical studies of religion tend to oversimplify the dynamic complexity of actual historical circumstances by presenting religious phenomena in a sequence of successive developments. A linear chronology does not do justice to the fact that older and innovative religious trends frequently share the same time and space. The arrangement of Parts 1 and 2 is thus intended to acknowledge the temporal overlap between the early expressions of Dizang worship, centered on the Scripture on the Ten Wheels, and their eventual extension.
74. The term is not intended to reect a kind of teleological view of historythat is, history on the way to some destined modernity. Nor is it intended to imply a western model that conceives of medieval as postclassical and prerenaissance.